We are giving Big Tech what they want: the strategic blunder of under-16 social media bans

We’ve known for fifteen years that individuals don’t need an account to be identifiable by social media companies (Garcia, 2012). Subsequent technological improvements now reveal highly accurate, intimate, real-time, dynamic information: tech companies know who you are, where you are, and what you are doing throughout most of the day (Clements & Horwood, 2024). Under-16 social media bans seem to encourage yet more access to governmental or biometric data and the use of analytic methods on children from companies that have already demonstrated untrustworthiness. This is startling enough, but what of the ways under-16 social media bans can serve big tech that we don’t yet seem to take into account?

Resource expenditure

My concern originates in the role bans can play in implementing effective long-term social and legal change: functioning as condition-setting mechanisms that suspend behaviour to enable larger comprehensive strategies, including new legislation and normative change, to gain traction.

A ban’s effectiveness doesn’t persist without these social (Bursztyn et al., 2026) and regulatory (Floridi, 2026) reinforcements. And once a ban has become demonstratively ineffective – once the ways to circumvent it have been identified and circulated, or compliance has become low – it’s challenging to reinstitute belief in its efficacy.

In other words, a social media ban is an exhaustible resource. Expending it without initiating long-term, large-scale educational and regulatory plans cannot succeed, undermines government credibility, and compromises future action (Floridi, 2026).

Setting and reinforcing norms

Governments clearly aren’t ready, but, once the condition setting mechanism of a ban is in place, anyone with the power to shape social norms alongside regulatory policies, as social media companies already can and do (Clements & Horwood, 2024), can capitalise on it: normalising public expectations to accept corporate powers of self-regulation (Floridi, 2026), invasive data collection, and privacy violations (De Cock, 2026). This suggests we have handed tech companies yet another opportunity to reshape our legal and social world.

For example, we are still reliant on these companies to run the checks, with limited insight into, or consequential power over, their analytic and algorithmic processes, or foundational design elements. Not only does this reinforce a sense of dependence on these companies to manage the back end of the ban, it offers the illusion of genuine regulatory action as a smokescreen for what are really more powers of self-regulation for companies: they retain management of platform access internally, acting as the policy enforcers themselves, while governments remain insufficiently equipped to tackle company behaviours (Reuters, 2026), and parents are inevitably left to flag ban failings, normalising a long-term trend that places accountability onto users (Penman & Turnbull, 2007). Meanwhile, the false impression adequate protections are in place alleviates public pressure for further regulation (Floridi, 2026).

Damaged trust in government

Worse, when it becomes clear that a strongly heralded action fails to deliver results, trust in government is damaged (Floridi, 2026: 10). For instance, in Australia only one in twenty under-16s voluntarily complies with the ban (Bursztyn et al., 2026), and just as the ban was being signed into law the eSafety commissioner reframed the policy, as “access delay”, softening the prohibitive language and permanence that leant the policy gravitas, even though Australia’s national dictionary had just elevated ‘social media ban’ as its word of the year (De Cock, 2026). Such contradiction undermines the sense of effectiveness, competence, and consistency necessary for effective regulation. Relatedly, such failings can also foster belief in the superior competence of private enterprise as better suited to governance, despite its self-serving agenda, while additionally coaxing governments to deploy neoliberal practices themselves (Morales et al., 2014), making for a concerning vulnerability in our democratic governance.

Children’s vulnerability and undermined democracy

We understand children’s vulnerable position as one that requires guardianship. But stewarding can be, in a tiny but significant step, re-engineered from caretaking to control: re-situating social beliefs about children’s culpability from innocent victims, to victims of their own behaviour, making them chief actors in their difficulties, thereby justifying the removal of their power.

This may be why, of the countries where bans are under consideration, they seem first to have appeared where the age of criminal responsibility has been flagged as concerningly low: 10 in much of Australia, and the UK. Sure, place children in adult prisons at 11, but God forbid they have a social media account! This undermines children’s protected status as somehow negotiable under the guise of ‘care’ or ‘community good’ to make them as litigiously responsible as adults. For example, bans degrade children’s “rights-holder” privilege of political speech (De Cock, 2026), as punishment for the behaviours (“addiction”, inappropriate access, etc.) that tech companies have themselves prompted by design (5Rights, 2021): tech companies behave poorly, children suffer suspension of their rights as though they are at fault.

Treating children this way opens the door on the sort of behaviours from private industry that are entirely unacceptable, and fortifies those already unfolding, while removing children’s ability to meaningfully object in important political spaces online: the theft of their personal data (and by extension, their own personhood [Floridi, 2005]), their perpetual surveillance (Boninger & Molnar, 2016), and the withholding of goods or services unless they surrender their personal information (e.g., transgressions unfolding in UK school environments that trade cafeteria access for biometrics like fingerprints [ICO, 2022]).

These experiences reshape the future social fabric by molding emerging young citizens’ expectations to accept invasive data collection for identity verification, perpetual surveillance, and the suspension of political expression as acceptable prices to pay for ‘safety’. Meanwhile, these same bans literally prevent schools from teaching students the tools to comprehend the dangers of their situation, or help them regain their agency. Indeed, what better way to ensure that the next generation of young people is unprepared for the machinations of big tech than to prohibit learning about it in hands-on, authentic ways while simultaneously institutionalising classroom data transgressions? And while these observable concerns focus on the Australian and British contexts, President von der Leyen’s recent EU address advocating similar EU policy, suggests more bans are forthcoming.

A strategic blunder

Cambridge researcher Amy Orben (2026) suggests the bans are a “gamble”, but a gamble implies at least a chance of success. Governments are now pinned in an impossible position: democratic practices like freedom of expression, or education, are pitted against other democratic principles like civic protection, fostering doubt in democratic legitimacy; tech companies occupy the drivers seat of policy and norm setting while we exhaust invaluable tools like bans; democracy and civil society are weakened, and children remain unprotected while tech companies’ mismanagement of our data, use of analytic manipulation, and wholesale extractive behaviours remain unaddressed. That’s not a gamble, it’s a strategic blunder, one demonstrating our ill-preparedness for the larger issues at play in our digital age civic world.

If bans are going to work in society’s favour, then strategy matters:

Timing: condition-setting mechanisms should be reserved for the moment governments are prepared to enact large-scale legal and educational initiatives.

Communication: language must accurately depict policy, and be deployed uniformly, to support a government’s perceived competence and capacity to deliver.

Consequences: As Luciano Floridi (2026) has recently argued, consequences must be credible, linked to norms, and enforceable by a concentrated, knowledgeable authority; approaches modelled in previous legislative successes like the EU’s GDPR or EU AI Act, (Floridi, 2026).

Design: anonymity, and the free speech of minors need to be protected through infrastructure that ensures protection for vulnerable actors by design (De Cock, 2026).

Most of all, we need long-term thinking, because today’s policies are tomorrow’s realities. We are weighing more than child safety in-the-moment. We must also consider child safety in the long-term, data companies’ progressive power over civic society, and the formation of young citizens’ democratic expectations.